792 



NATURE 



[February 17, 192 1 



the ship is at rest, or is carried uniformly forward 

 in a right line." 



Thus no experiment on board ship can ever 

 disclose the ship's velocity through the sea. The 

 matter stands differently to one who is free to 

 experiment with both the ship and the sea. Let 

 a sailor walk to the end of the bowsprit and drop 

 his lead into the sea. A circular ripple will spread 

 out; but every sailor knows that the point at 

 which his line enters the watei will not remain 

 at the centre of this circle. The velocity with 

 which the point of entry advances from the centre 

 of the circle will give the velocity of the ship 

 through the sea. 



If our earth is ploughing its way through a sea 

 of aether, an experiment conceived on similar 

 lines ought to reveal the velocity of the earth 

 through the jether. The famous Michelson-Morley 

 experiment was designed to this end. Our earth 

 was the ship ; the physical laboratory at Chicago 

 was the bowsprit. The dropping of the lead into 

 the sea was represented by the emission of a light- 

 signal, and the wave-front emanating from this 

 signal was the ripple on the sea of aether. In 

 the original experiments of Michelson and Morley 

 it was xiot possible to watch the progress of the 

 ripple directly, but sufficient information was ob- 

 tained by arranging mirrors to reflect the signal 

 back to the starting-point. In the recent experi- 

 ments of Majorana this difficulty is obviated, 

 although at the cost of some loss of refinement. 



From these and other experiments the result 

 invariably emerges that the wave-front appears to 

 be a sphere having the observer at its centre. 

 Thus on the hypothesis that our earth is sur- 

 rounded by a sea of aether, experiment shows that 

 the velocity of the earth relative to this sea of 

 aether is nil. We cannot suppose that the true 

 velocity is always ni7, for the earth is known to 

 be describing circles around the sun at a speed 

 of 30 km. a second, while the experiments were 

 sensitive enough to detect a velocity of one- 

 hundredth part of this. 



In view of the complete success which has at- 

 tended the hypothesis of relativity, it would scarcely 

 seem to be necessary to do more than mention 

 the various early hypotheses put forward to 

 account for these and similar experimental results. 

 Such were the hypotheses that the earth drags 

 the aether along with it (Arago, 1818) ; that matter 

 moving through the aether is contracted, as a 

 result of its motion, in just such a way as eter- 

 nally to conceal the earth's motion through the 

 aether from our measurements (FitzGerald, 1893 ; 

 Lorentz, 1895); and that light is a phenomenon of 

 corpuscular emission (Ritz). Each of these hypo- 

 theses explained some only of the facts to be ex- 

 plained, and failed with others. 



The theory of light has progressed largely 

 through the construction of mechanical models. 

 Every such model, if fruitful, suggests new laws 

 to be tested. So long as the laws suggested in 

 this way are confirmed by observation, the model 

 stands ; as soon as a predicted law is found to 

 fail, the model must be amended or abandoned. 

 NO. 2677, VOL. 106] 



Notable examples of such models have been the 

 corpuscular model of Newton, the elastic-solid 

 aether of Young and his followers, and the electro- 

 magnetic ffither of Faraday and Maxwell. The 

 first two of these have long ago served their 

 purpose and passed away. The time has now 

 clearly come when the last of these optical models, 

 the electromagnetic aether, must be either 

 amended or abandoned, and the indications are 

 strong that the less drastic course will not suffice. 



The construction of mechanical models is not, 

 however, the only known means of guidance to 

 the discovery of new laws of Nature. An even 

 more fruitful means of progress has been pro- 

 vided by tentative generalisation of known laws. 

 Proved laws a, b, c, d . . . are found to be special 

 cases of a more general law A, and the truth of A 

 is then seen to involve not only the detailed laws 

 a, b, c, d . . . which have led to A, but also other 

 detailed laws p, q, r, s . . . In this way we may 

 be guided to test the suggested new laws p, q, r, 

 s . . . , and the generalisation .\ is, of course, 

 strengthened or discredited according as p, q, r, s 

 . . . are confirmed by observation or not. Con- 

 spicuous instances of successful generalisations of 

 this kind are provided by the conservation of 

 energy and the second law of thermodynamics. 



Earlv in the present century Einstein and 

 Lorentz suggested a tentative generalisation of 

 this type, which is now known as the hyprothesis 

 of relativity. Since all experiments so far per- 

 formed had failed to disclose the velocity of the 

 earth through the assumed aether, it was natural 

 to generalise in the first place to the tentative 

 principle that this velocity could not, from the 

 nature of things, ever be revealed by any experi- 

 ment whatever. Generalised somewhat further by 

 the removal of the local reference to our earth, 

 the hypothesis assumed the form that all pheno- 

 mena of Nature are the same for an observer 

 moving with any uniform velocity as they are for 

 an observer at rest. This somewhat crude form 

 of statement shows that the hypothesis merely 

 generalises Newton's corollary V. quoted above, 

 so as to make it apply to all the phenomena of 

 Nature. Since, however, the acceptance of the 

 hypothesis makes it impossible to define what is 

 meant by a state of rest, it is better to express the 

 hypothesis in the form that all the phenomena of 

 Nature are the same for any two observers who 

 move relative to one another with a constant velo- 

 city. 



This hypothesis is known already to be true as 

 regards the mechanical forces considered in New- 

 ton's laws. Naturally, also, it is true as regards 

 the optical phenomena investigated in the 

 Michelson-Morley and similar experiments, for it 

 is out of these phenomena that the hypothesis 

 arose. The crucial test occurs when laws in other 

 fields of science are deduced from the hypothesis 

 and compared with observation. The hypothesis 

 has been very thoroughly tested in the field of 

 electromagnetism, and in every single case has 

 emerged triumphant, .^s conspicuous instances of 

 its success may be mentioned : The explanation 



